Main content

Enterprise hits and misses - ChatGPT-4o gets an early enterprise grade, Meta bails on Workplace, and AI forces a skills rethink

Jon Reed Profile picture for user jreed May 20, 2024
Summary:
This week - Meta bails on Workplace, but was this a serious enterprise play? OpenAI releases ChatGPT-4o - should the enterprise care? AI is forcing a skills rethink, but will it usurp developers? As always, your whiffs.

building-blocks-fall

Lead story - ChatGPT-4o - a tiny step for AGI, a giant leap for emotional (like) AI?

I couldn't resist the snark of putting a question mark at the end of George's post. GPT-4o is here, and it deserves a thorough review - will it aid the enterprise? 

George has written some of the best enterprise AI posts on diginomica this year, always balancing the potentials and the pitfalls. But I get the sense he might be more bullish on GPT-4o than I am. On the emotional intelligence of GPT-4o, George writes about the implications of "audio as a first-class citizen in generating responses," cross-referenced against other modalities where needed: 

The upshot is that it will make standardizing the analysis of conversational cues across different speakers and cultures easier.  Early use cases might include better customer service, sales, and counselor coaching tools. It may be tempting to roll this out for customer-facing chatbots to improve user experience. However, this could also introduce new reputational or business risks.

OpenAI plans on offering a free version of some of its latest models. George issues a caveat: 

But as with all free products, it's worth cautioning that when big tech companies offer us a service for free, the product ends up being you. With social media, that means monetizing our attention. 

Free versions of this type come off as a consumer play, not an enterprise one. But as George points out, there is enterprise relevance here too: better user experience. 

 Once the service goes live, it will definitely be worth exploring to help appreciate how it and future competitive offerings may shape and improve user experience design. 

I view emotional intelligence as a low bar for AI. Why? Because humans are more than willing to project emotions and understanding onto objects they want to relate to, no matter how flawed their grasp is (I even do that with Alexa, and Alexa has no insight into my comings and goings, even after six years of exchanging verbiage, and has yet to propose a product I actually need). I don't view OpenAI as the most important company when it comes to enterprise AI - nor do I see OpenAI as the leader in any pursuit of AGI; they are too busy/under too much investor pressure to monetize, rather than to hit the white board, which is the toll the real pursuit of AGI requires. 

But here's an interesting one: with its audio emphasis, will GPT-4o supplant the mediocre mobile voice assistants (Siri et al)? The bar is clunky low here, and OpenAI excels at more entertaining interactions, even when accuracy falls short. 

On the enterprise side, George is onto something with the multi-modal, multi-lingual aspect. OpenAI isn't going to win awards for cross-cultural sensitivity any time soon, but translations and multi-lingual dexterity might be the top use case here for enterprises, based on the release reviews I've seen. But, limitations abound: GPT-4o’s Chinese token-training data is polluted by spam and porn websites

As for a reduction in hallucinations and operational costs, these are incremental improvements at best. Meanwhile, enterprises are figuring out how to reduce AI costs by reducing dependence on models trained on the known universe when that's overkill to their use case. The ROI of gen AI will remain challenging, due to everything from operating costs to the need for human review of (most) output, which impacts both the productivity and the financial return. Still, OpenAI is the elephant in the room, the biggest reason we are grappling with this, and, without doubt, the best at engaging users in appealing ways. That's one thing enterprises could stand to learn from. 

Diginomica picks - my top stories on diginomica this week

Vendor analysis, diginomica style. Here's my top choices from our vendor coverage:

Hit the tarmac with diginomica - as spring event season rolls on. 

Boomi World 2024 - Sarah was on the ground in Denver, and posted a flurry of use cases and updates, including: Boomi World 2024 - AI futures according to Boomi and Red Hat leaders and Boomi World 2024 - why organizations need to tackle data fragmentation more than ever if the benefits of AI are to be captured

Planful Perform 2024 - Brian and I made the trek to sunny San Diego last week. My roundup of the underrated news stories is up: Planful Perform 2024 - customers kick tires on AI, but will financial planning break down the decision silos? I've also posted the podcast wrap recorded with Brian: Planful Perform ’24 Review - impact of AI on finance and customer priorities. Plus: a bonus short podcast, ESG for finance - where do we stand now?

A few more vendor picks, without the quotables:

Jon's grab bag - Madeline gets it done again with Three hundred meetings with young men in gilets and chinos who “laughed in my face” – how Anne Boden founded Starling Bank. Martin deciphers buzzword bingo in The art of super-opting - before you monetize data, ‘contractize’ it.  Finally, Stuart examines a high stakes industry for AI: education: "Getting past the plagiarism paranoia and the resultant negativity is going to be critical. Academic circles need to learn - or be taught - the wider benefits potential of AI in education." (EducAItion in action - two US academic institutions offer lessons on AI and education beyond plagiarism paranoia)

Speaking of Alexa, catch diginomica on Alexa - I'll have more on this in the next couple weeks, but if you're using Alexa, consider adding diginomica as a general Alexa skill. We've already had diginomica news summaries available in the Flash Briefings, but now it's a general skill. Once you add it, ask Alexa to "open enterprise news," and you can get the latest from diginomica - with no GPT4o needed...

Best of the enterprise web

Waiter suggesting a bottle of wine to a customer

My top seven

Overworked businessman

Whiffs

404 Media reports that AI Generated Hentai Is Viral All Over Facebook (click at your own risque). Meanwhile, Facebook deleted a news article about AI-generated porn. Oh well - why read about it when you can experience it? 

Then there was this: 

The lesson isn't the whiff; it's the field-tested rationale for multi-cloud... Amazon is trying to keep workers cool, but they aren't impressed: 

Finally, Frank Scavo gave gen AI a shot... for stand up comedy: 

Humorous fortune cookies turn out to be a higher bar than summarizing a dense white paper, or a long-@ss meeting! Go figure...

If you find an #ensw piece that qualifies for hits and misses - in a good or bad way - let me know in the comments as Clive (almost) always does. Most Enterprise hits and misses articles are selected from my curated @jonerpnewsfeed.

Loading
A grey colored placeholder image